The prevalence of type 2 diabetes (DM) in older persons is increasing rapidly. DM increases the risk for Mild Cognitive Impairment (MCI), which is a transition state between normal cognition and dementia that is often characterized by memory and executive function deficits. These deficits reduce adherence to DM medications, which worsens glycemic control and increases the risk for adverse DM-related health outcomes. Improving medication adherence may prevent these outcomes and reduce health care costs. This is important to all older persons with DM but particularly to older African Americans (AAs). They have twice the rate of DM, worse cognitive function, lower medication adherence, and worse glycemic control than whites. One million older AAs now have DM and their number will double by 2030. Because 30% also have MCI, low medication adherence is an important problem for them. This necessitates culturally relevant interventions that compensate for their cognitive deficits and improves their medication adherence and glycemic control. We propose a randomized controlled clinical trial to test the efficacy of a collaborative Primary Care-Occupational Therapy (PC-OT) intervention to lower hemoglobin A1c (HbA1c) levels in older AAs with DM, MCI, HbA1c e 7.5%, and d 80% adherence to an oral hypoglycemic medication. PC-OT consists of: 1) primary care physician (PCP) - occupational therapist (OT) collaboration; 2) DM education tailored to cognitive impairment; 3) in-home OT cognitive-functional assessment; and 4) OT-delivered Behavior Activation to increase adherence to medications and other diabetes self-management (DSM) practices (e.g., diet). We will recruit 100 participants from primary care clinics and randomize them to PC-OT or Enhanced Usual Care (EUC). EUC is usual medical care plus low intensity DM education delivered by community health workers. Participants in both PC- OT and EUC will have 6 initial in-home treatment sessions over 3 months, and then 3 booster sessions during this 12 month study. The primary outcome is a reduction in HbA1c of 0.5%, which reduces the risk of adverse medical events. The primary efficacy analysis compares the proportion of participants in PC-OT and EUC who achieve this outcome at month 6 (short term effect) and at month 12 (maintenance effect). We will measure medication adherence using an electronic Medication Event Monitoring System, prescription refills, and self- reports. A secondary aim determines if improving medication adherence mediates PC-OT's impact on HbA1c levels. We will also evaluate PC-OT's effect on other DSM practices; ER visits and hospitalizations; cognition; function; mood; and quality of life; and PC-OT's costs and net financial benefits. This is the first study to determine if PCPs, collaborating with OTs (who are experts in developing strategies to compensate for cognitive/physical deficits), can improve medication adherence and glycemic control, and prevent cognitive and functional decline in older persons with DM and MCI. If PC-OT is effective in a high risk population of older AAs, its benefits may extend to all older persons with DM and have enormous public health significance.